Natural Knowledge in the Classical World 43idea repeatedly in his cosmological works. We have already seen the ar-
gument that eternal and continuous motion is necessarily circular, and
Aristotle showed in On the Heavens that the cosmos is in motion eter-
nally and continuously. There are also, as we shall see in a later section,
powerful theological considerations that further compelled Aristotle to
these conclusions.
By Pliny’s day, Aristotle’s eternal, spherical cosmos had been subjected
to much scrutiny and revision. One school that was very prominent from
the third century BCE and for half a millennium afterward, that of the
Stoics, accepted the sphericity and divinity of the cosmos, as did many of
their contemporary Platonists. The Stoics, however, offered a variation on
its eternity, arguing that the cosmos was subject to periodic confl agration
and rebirth following a strict chain of natural causes. The matter in the
cosmos persisted through the destruction (how could it do otherwise—
how could something become nothing?). Because there were no random
events but only cause following cause in a strict chain, the Stoics argued
that the reordering of the cosmos, its birth after any one confl agration,
was necessarily identical to its reordering after any other confl agration.
This meant that the unfolding of the universe, all its causes and all its
events, happened in exactly the same way again and again, a fi nite period
of cosmic history recurring identically to eternity—the eternal recurrence
of the same, to use Nietzsche’s wording for the idea he borrowed from
them. What is interesting here is that the eternal recurrence—and with it
a complex fatalism—emerges from a belief in strict causal necessity, that
identical matter subjected to identical causes must always react in identi-
cal ways.^15 The periodicity further tied in neatly with contemporary astro-
nomical contexts where cosmic periodicity, the eventual return of planets
to identical confi gurations, often was seen to imply an astrological return
to identical cosmic infl uences and effects.
But one school rejected all versions of the fi nite world in favor of an
infi nite cosmos with no inhering or interfering divinity whatsoever. Dat-
ing back to before Aristotle’s time, several versions of atomism had been
proposed, with Epicurean atomism eventually taking the predominant
place from the third century BCE onward. Atomists argued that motion
would be impossible in a universe that was completely full of matter, and
so there must be void spaces between things if anything were to move at
all. Furthermore, the particles of matter would have to be perfectly indivis-
ible (in Greek a- tomos), in order to explain the existence of any perceivable
objects. The argument was that destruction happens faster than construc-
tion, and unless there were a stopping point—indivisibility—for destruc-
tion, there could be no macroscopic cohesions of atoms for us to observe.